Tibo Questions Companies Using Opus Over Sol
Replies cite task failures, context limits, and compliance needs when explaining model choices.
Tibo asked companies why they still use Opus instead of Sol when price is a factor. Rylan Schaeffer replied that Sol 5.6 often runs for days without finishing tasks that Fable then completes. Alex Cohen said his team would switch to ChatGPT if a BAA were offered on Pro plans. Multiple developers noted they choose Fable for its full 1M context window, which they say GPT fails to deliver reliably due to compaction or OAuth limits.
If you have a company and still use Opus instead of Sol, why so? Does price not matter to you? What's something that would convince you to switch?
Combined views
1.3M
7 posts, first seen 8d ago
Reactions from ranked influencers
6 posts@thsottiaux Sol 5.6 will work for days and not get the job done. I hand it to Fable and Fable finishes it. Has happened a few times now.
@thsottiaux I use Opus because it's all I'm allowed to use for the other 50% of my Claude Code sub. I use it for tedious work when I have enough time and pain tolerance to spare.
@thsottiaux Because GPT's context sucks, I hate compaction, I do work that requires the full 1M ctx of Fable (not Opus, lol) and the model is just not as good
@thsottiaux If yall would sign a BAA for Pro plans we’d have everyone switch to ChatGPT/Codex ASAP.
actually yeah, the weird thing is I actually would use openai via oauth more, but for some reason they never even fixed the bug that makes oauth not have 1mm context access. and kimi k3 is way cheaper and with clean 1MM context, so for peak I use fable and for cheap i use kimi.…
@Teknium @thsottiaux (hermes w nous portal goated btw)